• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Suggestions for causes of games crashing since RX580 fitted

Soldato
Joined
27 Dec 2009
Posts
2,727
Location
Gillingham, Kent
Hi everyone. A good friend of mine treated himself to an RX580 for Christmas after his trusty HD7950 was struggling too much in modern games. He also treated himself to an SSD (two actually, one old one for OS and one new one for games) and a fresh Windows 10 install rather than his many-year-old-upgraded-from-Win-7-to-10 OS.

Since the change over, he's had issues with games crashing to desktop or freezing and I'm struggling to come up with suggestions for him to try.

His spec is:
i7-2600K (running at stock with standard Intel heatsink)
Gigabyte P67A-D3-B3
16GB DDR3 1600 (4x4GB)
Kingston SSDNow V100 128GB (OS)
Samsung 850 EVO 250GB (Games)
Western Digital WD10EALX (More games)
Powercolor RX580 8GB Red Dragon
Coolermaster Silent Pro 700W

He's having consistent issues with PUBG and Wolfenstein II crashing after 5-20 minutes, generally with a crash to desktop but the game still shows as running in the background - but frozen and requires killing in Task Manager.

I suspected the power supply as barring the SSDs and the GPU, the system has been running fine for years with the 7950 in, so I wondered if the RX580 was putting more strain on parts of the PSU than the 7950 had. I took my Corsair TX650M from one of my PCs (it usually runs an i7-2600K, SSD, HDD and R9 290 Tri-X faultlessly, so figured it'd be fine with his PC) and he's tried it in his PC, but with the same games crashing after up to 20 minutes of gameplay.

He's also tried the last 3 AMD drivers, but again with the same issue.

Anyone got anything we can try to narrow this down, or run into the issue themselves? At the moment, I'm thinking of testing the GPU in one of my PCs, but we live almost an hour apart and both work busy schedules, so getting time in the same room is tricky, so any suggestions he can try are gratefully received. Cheers.
 
Running any temp monitoring software?
Is it only these 2 games he's having issues with. Can't speak for wolfenstein but pubg can be damn erratic!
Ran memtest to rule out dodgy ram?
Intel stock coolers are pants even for non oc'd chips. I'd be looking at temps initially. The 580 could be dumping more heat into the case and causing issues.
 
I had an RX580 prebuilt system and it would hard lock all the time. I had to send it back because it couldn't last 5 minutes in a game. If I was him I would send it back.
 
When is the last time he reinstalled Windows?

I had problems with similar crashes on my 480 which I could help slightly by increasing voltage, disabling overlays and messing with various combinations of vsync settings but they were finally stopped when I reinstalled windows after the recent fall creators update.

No amount of DDU ing and different drivers cured it.

Giz
 
And if the 580 isn't directly at fault?

Then the shop test it and tell you so.

Otherwise you get a replacement and haven't had to waste time from your "busy" schedule only to find out its probably duff anyway. The one I had did the same thing as OP and at time when trying to diagnose mine I read loads of similar issues from people.

The one I had was an Asus if that helps any.
 
Good suggestions everyone, thank you. I'll ask him to have a look at CPU temperatures, run Memtest for a while, and I'll raise the idea of a Windows reinstall (maybe something is broken in this one, even if it's new?). I'm happy to send the card back if it's faulty, obviously, but just would like to make myself 100% sure it's the card before starting the RMA process, if possible.
 
Then the shop test it and tell you so.

Otherwise you get a replacement and haven't had to waste time from your "busy" schedule only to find out its probably duff anyway. The one I had did the same thing as OP and at time when trying to diagnose mine I read loads of similar issues from people.

The one I had was an Asus if that helps any.
Sorry I replied as I did as it seemed you were suggesting to simply return it without testing/checking to see if it was the fault. If you send something back under rma and it turns out not to be faulty you could be looking a testing fees and return postage plus the postage you paid to send it back in the first place. Not to mention the time you'll be without a gpu.
Taking an hour or 2, at worst, to run a few checks is worth wasting out of my busy schedule. I'd hate having a schedule that was so busy I couldn't spare an hour or 2, must be a horrible way to live :p
 
Try upping the GPU Power Limit to +50% in Wattman, click Apply and test.

If crashing continues, try upping the voltage in Wattman State 7 to 1200 and click Apply and test. If this fixes it, the GPU is unstable at default settings, but the voltage bump makes it stable.

If the crashing continues, try DDU and an older driver, something from 3-6 months ago and test.

If the crashing continues, can you try it in your Motherboard/system?

If the crashing continues, the GPU is most likely faulty.
 
Thanks very much again for the suggestions everyone. He's tried several things over the past few days, and (touch wood) increasing the GPU power limit by 50% appears to do the trick.

He's back to being a happy bunny with his new graphics card now. :)
 
Gah! Spoke too soon. :(

He's just reported that PUBG now runs for 2-3 games before crashing, and Wolfenstein can run for 60-90 minutes before crashing, but the games are still crashing. I'm unwilling to ask him to start messing around with voltages, so I'm going to ask OcUK if they'll authorise an RMA because the card isn't stable out of the box. Hopefully they say yes and he ends up with a card that does work without tweaking.

Thanks again for the help everyone.
 
Gah! Spoke too soon. :(

He's just reported that PUBG now runs for 2-3 games before crashing, and Wolfenstein can run for 60-90 minutes before crashing, but the games are still crashing. I'm unwilling to ask him to start messing around with voltages, so I'm going to ask OcUK if they'll authorise an RMA because the card isn't stable out of the box. Hopefully they say yes and he ends up with a card that does work without tweaking.

Thanks again for the help everyone.


PUBG was the game where mine would crash to a grey screen. I'd get it sent back. You shouldn't need to be tweaking a new card so much to get it working.
 
Back
Top Bottom